Your Company Has 6 Front Doors (And Employees Hate All of Them)

Your company doesn't have one front door — it has six. IT has one. HR has one. Facilities has one. Legal ops has one. And employees hate every single one of them.

In this episode, Pops and Steele pop the hood on ServiceNow Employee Center — not as another tech deployment, but as an enterprise transformation play. They tackle the governance question every organization avoids (centralized vs. federated vs. product-led ownership), why portal fragmentation is a measurable business risk and not just a UX nitpick, and how to build an executive case using real pain points and testimonials instead of another feature-list slide.

The back half gets tactical: who owns which KPI, chargeback vs. showback cost models, and the cadence for reviewing metrics with stakeholders — weekly at the platform level, monthly red/yellow/green rollups to execs, and quarterly strategic moves through CAB.

A single employee front door isn't a technology rollout — it's an organizational commitment. Product-led ownership with domain accountability is the "Goldilocks" governance model: centralized ownership becomes a bottleneck at scale, federated governance drifts without a brand standard, but product-led ownership with strong stakeholder engagement holds the line while still giving each domain room to serve its users well.

If you're an IT leader, ServiceNow product owner, or anyone fighting portal sprawl inside a large or regulated organization (yes, including law firms), this one's for you.

🍺 New episodes every week 

🚗 Subscribe for more under-the-hood conversations on ServiceNow, AI, and IT leadership.

ServiceNow Employee Center, Employee Experience, Portal Fragmentation, Business Transformation, Governance Model, Product-Led Ownership, Federated Governance, Shadow IT, Chargeback vs Showback, KPI Ownership, Executive Sponsorship, CAB, OCM (Organizational Change Management), Digital Front Door, ROI Metrics, Legal Industry IT

If you're staring down your own portal sprawl, drop a comment — how many front doors does your org actually have? Subscribe for more under-the-hood conversations on ServiceNow, AI, and IT leadership, and we'll see ya later.

Transcript

SPEAKER_00 0:00

So how do you quantify the economic impact of portal fragmentation uh using metrics such as time to resolution, ticket deflection, and employee productivity loss?

SPEAKER_01 0:16

So the first thing I think of it just as one of your friction sale points in the in going with a unified employee center in a sense, is it's not about resolution time only. That's not your cost factor. It's everything before the ticket. When the employee can't find the right portal, they can't they submit to the wrong team, it gets rerouted, they play volleyball with it, where they they give up and they can call the help, you know, they don't call the help desk directly. All those costs you've lost already.

SPEAKER_00 0:48

Yeah, I I love that. That's uh was not even thinking about all the things that happened before that ticket submitted. Uh that is a great point. And uh yeah, I I think it's possible to assign numbers to those activities, track the metric, and then just do the math on the same.

SPEAKER_01 2:02

That's what's happening now. It's not lava inside of a lot of organizations. Sharp builds the portal, IT builds a portal, Lego builds a portal, and facilities has a form buried somewhere on the internet. So in your employees, they're just out there guessing where to go. So today we're trying to pull this into the garage and pop the hood on ServiceNow Employees Center. Not as a tech tool, but as a business transformation play. We're talking about what it actually takes to go from fragmented siloed service experiences to a single unified front door that employees just actually trust and use. So this isn't only a usability conversation, though. Portal fragmentation is measure is a measurable business risk. It costs time, it costs money. In environments like legal or healthcare, it can cost you in compliance and shadow ID. So whether you're the one trying to make the case to leadership, the one building the architecture, or the one caught in the middle of organizational policies, this episode is probably for you. So Steele and I are going to walk through the strategy, the governance, the AI angle with analysis, the migration realities, and yes, how you can actually prove ROI when you're done. So, Steele, let's get into it.

SPEAKER_00 3:10

Do it. The front door of any business is important. If the eyes are the window to the soul, the employee front door is the soul of your associate experience. So let's start with strategy and ownership. What operating model best sustains a single front door over time? Centralized ownership, federated governance, or a product-led model with domain accountability?

SPEAKER_01 3:37

Go. I would say none of them work in isolation.

SPEAKER_00 3:40

You know, by that. Yeah, I think that's a a trick question.

SPEAKER_01 3:45

It is. But I think the product-led model, if you have domain accountabilities, um, is probably the one that survives the longest if you had to choose one. Centralized feels clean, you know, it sounds right, but in practice just becomes a bottleneck, and everyone seems to be waiting on IT or something like that. So federated sounds like a mature way of doing it. I mean, when I say it's federated, I it sounds pretty cool, pretty thick. But even that federation, I think, drifts.

SPEAKER_00 4:14

Ian, I I'd agree with all three of those assessments there. I think if you're really small, you can get away with centralized ownership because you can still move at some sort of agile speed. Once you grow larger, that becomes much more difficult. You start seeing the cycle time slow down.

unknown 4:32

Yeah.

SPEAKER_00 4:34

For federated governance, I think drift is definitely the issue. You'll have folks who might be given a certain space and build a certain way. It's important to make sure that they're following some kind of brand plan or brand standard. And I think the Goldilocks method is that product-led model with domain accountability, in which case you have somebody who's thinking about it as a product and driving towards that vision of the single front door, kind of being the keeper of that while still having engagement with stakeholders and process experts for their individual domains so that you have a consistent experience, but also you have the best experience that you can inside of each of those domains. How do you position Employee Center as a business transformation initiative rather than a ServiceNow project to secure cross-functional buy-in?

SPEAKER_01 5:29

Oh my gosh. So I think you're saying something like we're redesigning how our people find requests, receive services across the company, and we're using technology to make that seamless. I definitely think people can argue that that's one-sided or not.

SPEAKER_00 5:45

I think that's part of it. I think the I love starting with pinpoints. And when you look at how many tools are out there across the organization to try to accomplish these tasks, do you have an internet that has knowledge or outdated knowledge? Do you have another portal out there that has the little square boxes that links to all the other applications in the organization? Are those interoperable? Do you have a lot of integrations to consider? And when you actually map out what that architecture looks like for all those solutions, it's it's it'd probably surprise you. It's pretty, it can be pretty vast. And bringing all of that together in a cohesive strategy, reducing that barrier to get things done, reducing how painful it is to be an employee in your own company, I think really sells that vision because at the end of the day, everybody is just trying to get work done. Nobody wants to go to a portal because for for technology's sake, and I that's where this excels is it's connected to everything. You have all your data, you have all your integrations, you can just go. Yeah, you're right.

SPEAKER_01 6:50

I mean, you meet with friction, and you meet with here's what we're solving. These are all the issues we're having today. And you have testimony from various departments and teams speaking to those friction, those post-pain areas. And then you have representatives from those various teams that are like your flag carriers. You know, I'm always big on that. I love having testimony from any outside. You know, I I think I think that's probably your biggest win is to make sure you have support outside of IT, you know, from different departments and teams and they understand what you're trying to accomplish.

SPEAKER_00 7:22

I I think all of that just helps build that business case. And when you go through and you perform those exercises, you get the testimonials. That business case, I want to say it builds itself. But if you are able to just demonstrate that vision in a demo, it sells it.

SPEAKER_01 7:37

Yeah, I think you opposite of sometimes you talk about tech projects, you talk about outcomes, not capabilities. Yeah, don't talk about here's all the whiz bang whiz bang things you can do. You say you talk more about look at all the solutions this provides or how to save time or reduce this or you know, made it easier, simpler here, you know, things like that. I think that's where you have to speak. You can't talk about capabilities and you know, tooling.

SPEAKER_00 8:02

Absolutely agree.

SPEAKER_01 8:03

All right. So, where where does accountability for employee experience KPIs live? And how do you prevent the diffusion of like ownership across IT and business units?

SPEAKER_00 8:14

No, just blame your manager. It works every time. No. Oh, I'm kidding though. Uh so I think there is a portion of that, which does belong with leadership, tying those KPIs to OKRs, some kind of key result that you're trying to accomplish that is owned by an executive sponsor if it is a project. Uh, if it's not a project, then essentially by that sponsoring organization. I think there is various different levels, especially if you think about the product-led model. You've got product owners, you have product managers, then you have leadership at various different levels. And so I think on the associate experience side, that filters down. But when you think about those outcomes and how we're achieving against our strategic goals, enterprise front door goals, then that filters up to leadership. Uh I think if you don't have the right level of buy-in from the leadership level, or you don't have the right level of tracking at the execution layer, then it falls apart. And then you don't get the value or execute against outcomes, or you just don't know where you are.

SPEAKER_01 9:17

Okay. So what is your executive level narrative for why fragmentation is a measurable business risk, not just the usability issue?

SPEAKER_00 9:25

I think there's several different ways to approach that. And it is going to depend on what really resonates when you're with your leadership. It's a very diplomatic answer. And if you consider it from a compliance lens, what does it take to secure all those entry points, all those attack, the attack surfaces, vectors, security stakeholder? If you have somebody who writes the checks, how much does it cost to support all those different tools? Do we really need all those different tools? Uh, from an enterprise architecture perspective, how many boundary systems are we having to support and integrate with? What is the tax on IT? Uh, how does that slow down business cycles? And from the business perspective, if we want to deliver quick, managing all of this takes our cycles. We we can't be as agile as we could if we had less of a surface to support. So, which kind of shifting gears a little bit, still very much part of this strategic conversation around the organizational policies and also the dynamics of the politics around this conversation. So, which stakeholders have the most to lose in consolidation? And how do you design incentives that align them to a shared outcome? I think those are probably on paper.

SPEAKER_01 10:39

The biggest losers are the people who run each portal, you know, who built equity in it, who probably stood behind it, led a campaign of their own, you know, for that portal. Those are some of the battles you'll have. And it's like it's not that your portal doesn't work today as as much as it isn't as tied in and effective with everything else today, with how much data is integrated and that beehive connection of data. You know, it's gotta be sold that way.

SPEAKER_00 11:10

Yeah, absolutely agree. It's that concept of it's a point solution. I'm going over here to this neighborhood for one thing. I'm going over here to this neighborhood for one thing. Where's my Dior Duche? Can I can I just door dash it and get it provided to me?

SPEAKER_01 11:26

Yeah, and the other hidden thread behind that is budgeting and really headcount. Are you reducing headcount? Is there a battle battle to be brought there against you because you're cutting people in somebody's team?

SPEAKER_00 11:38

Yeah, then you you gotta figure out if the the human capital part of that conversation. Are you able to shift resources? Can you use them somewhere else? I think in the current climate, kind of the play I would try and make is reduce the amount of tools so that you could shift some of those support folks over to learning AI, building skills and process improvement via AI that would then directly feed back into that portal, probably upping your knowledge management game would be the first fixing data.

SPEAKER_01 12:08

Yeah, someone once told me give them a seat, not a grave.

SPEAKER_00 12:12

Wow. That sounds like a bumper sticker if I've ever heard one. No context either. Give them a seat, not a grave.

SPEAKER_01 12:19

I mean, because there's a lot of battles when you get it. I mean, we went into them with I don't know, consolidating, going from several data centers in each office over at our former firm to centralized data centers. Still, you're changing people's jobs, they worry about what that change means to them, how that affects them, where their trust goes, are you making your days worse? You know, so make give them a seat in this, you know.

SPEAKER_00 12:42

And on the plus side, they can now become cloud engineers. So there's there's the upskilling in that.

SPEAKER_01 12:48

Then they can go back to physical engineers that later.

SPEAKER_00 12:51

Yeah, full circle. It was popular in the 70s, it's coming back. So, how do you address brand identity concerns from departments that built their own existing portals? So, if you've got those folks that were managing that infrastructure, providing that experience, maybe they had white glove service, and now you're saying that's going away. How do you address those concerns? I I feel like this is a uh the this whole episode is like stomp the chump or something.

SPEAKER_01 13:20

It is. These are these are tough. And and for anybody who's listening and knows maybe where I've worked or who I work for or whatever, these answers are a mix of my career, not a certain location. So they can't take an oh, I know when that happened. That's not gonna be the case. I think let me see the what the question is again. How do you address brand and identity? First brand, brand has got to be your firm brand. You gotta try to match the firm brand. You you want to make that be consistent and cohesive, sort of with your website, with marketing. You have them on board. You don't you don't need another you don't need another enemy on this fight. That why did you pick blue and purple and green and yellow, whatever those things are. So brand, I guess if we speak of it that way, but identity is something I get. I think one thing you gotta say is this is the front door to the establishment. It's not the identity of this of each establishment. So I think that's one thing. I think identity doesn't need to be the skeleton or the framing, it can be the content. So if you built like actions, systems, menuing, sort of those sort of things, but the content fits the team that is supporting content, meaning they have their own choices and options and values and a lot of it's more than more than just like text. It I think the content be theirs, the framing, the skeleton, uh, how it maneuvers, it it belongs to the team, belongs to everyone, you know. I think that's a point.

SPEAKER_00 14:49

I I think if you take that product-led approach and you get you you kind of send out your stakeholder surveys and you figure out who's got those concerns, bring them in early because they're gonna be the loudest attractors, but they're all they they'll also will give you the best feedback uh that you can use. And those will be the folks that you'll make sure that they're involved, their skin in the game. And if you have a set dev program or they just have that close partnership with the product manager for that space, that could help address a lot of concerns because then they don't feel like they're being ignored. I I think that's at least part of it is that they know that they're being heard and that their feedback is not going into a black box that nobody's looking at. I think that's a big one. I think a lot of folks try to do the best that they can for the business. And if they feel like that is an intent, you're transparent. I think that's how you win some friends there.

SPEAKER_02 15:46

Yeah, I agree.

SPEAKER_01 15:47

So, what governance mechanisms prevent like regression into the silos again after initial rollout?

SPEAKER_00 15:54

Yeah, maybe I'll I'll do bonus points and add it for each of the operating models. I think that this is tough, but I think it goes back to whether you're centralized, product led, or federated. This goes back to that buy-in around the KPIs. You know, you have your outcomes that you're trying to achieve. You have KPIs associated to OKRs. Are we executing on the operational layer? Are tickets being created? Are they being closed? Did our deflection rates go up? Did our MTTR go down? Is it providing business value? And then how are we OCMing that or organizational change management? How are we sharing that story out? Are we sharing those wins? Are we still creating excitement even after we've gone live? And then are we still there helping to educate users both when new features are rolled out, but then do we just have a training to make sure that folks who come on board know how to use the portal and the features that are there to help them? I think the ones generally work across the board. It just takes consistency.

SPEAKER_01 16:54

So so how do you handle the exceptions? Those teams that legitimately maybe require distinct experiences due to regulatory.

SPEAKER_00 17:02

Yeah, and certainly there are some cases. HR and compliance is a really good one. Uh, I've seen some very interesting requirements in those spaces. And I think that goes back to having that tight relationship with the product manager. Not necessarily that the stakeholder has the tight relationship, but that that product manager has a good relationship with the product owner and that there's architecture in there and kind of as a trifecta or team, looking at what's fit for the platform, how can we solve this or solution this? And just making sure that we're considering all our options before we go through, we customize or we create something crazy. So that's that's the way that I think about it.

SPEAKER_01 17:45

So some of that isn't that just also a lot of it's usually like HR. You're you want to regulate the content. I mean, regulate how the questions can be asked or what can be retrieved. You know, this I guess there's some and you don't you want that kind of security throughout? Because as as spread out and tied as everything is, you don't want anything leaking through. You know what I mean?

SPEAKER_00 18:10

Yeah. Um I I also think that um have a prioritization mechanism. Yeah, you have intake, you have prioritization, and then you've got that solution review and some kind of technical governance so that you're prepared to handle uh either we're we're gonna have to, we legitimately have to have an exception and do something different versus we don't need to do it this way. We could solve it a different way, versus actually there is something that's already there for you. Sometimes they'll be on the roadmap and you just gotta raise it in priority.

SPEAKER_01 18:42

So how do you design okay? Now we're switching a little bit to some industry specific, but we'll we don't have to go really industry specific, but we may. But how do you design a unified portal respect, strict data segregation, confidentiality, and ethical boundaries?

SPEAKER_00 18:57

You wave a magic bond and you say gone. No.

SPEAKER_02 19:01

Yeah.

SPEAKER_00 19:02

I think this goes back to just some solid security principles and platform capabilities. Designing everything around RBAC seems to work pretty well. At least that is the current school of thought, and I think has been for the past decade plus maybe longer, where the person who is accessing the portal has a business reason for that access, is provided access commensurate to what they're trying to access, and they can only access data that they should have access to because of the roles that they have. I think that works in kind of a single tenant when it's your own company. If you're an MSP and you provide that instance to multiple customers, then I think you get into domain separation and taking a lot of Tylenol.

SPEAKER_01 19:50

Which is actually a little bit easier in a sense than having to separate your own departments and company that the same company that you know ties a lot together.

unknown 19:59

Yeah.

SPEAKER_00 20:06

Yeah, it's fenced off. There are there's some overhead that that adds and some complexity around access controls, but there are tools for handling those different scenarios on the platform to make sure that you are secure from a data perspective, compliance perspective. On the ethical side, I think that goes back to requirements and probably local laws. So those are also good things to gather requirements for when you're you're starting out, but then also keep tabs of even after you've gone live. I'm sure you're all familiar with GDPR, California's privacy laws. I can't remember what the specific one do I have to help me remember that one. You got lots of privacy controls, lots of different requirements. It's a great day when everybody says, hey, you know what, this is good. We'll all agree, and then we can just implement one control that satisfies everybody in an ideal world, but get those requirements early. So what architectural patterns? Well, I think we just kind of answered it in the last one, but what architectural patterns support role-based experiences for attorneys, staff, and partners without fragmenting the platform?

SPEAKER_01 21:17

Yeah, I think again, following the what one platform many lenses approach, it's uh taxonomy first. So instead of worrying about the theme of the widgets, you worry about what topics show up, what catalog items are visible, what announcements are rendered, just things like that. You know, how attorneys shouldn't see IT asset disposal requests, you know, just things like that, right? The next one, um thinking of it as an not as an attorney portal or staff portal, but as topical domains, like legal technology, HR services, finance facilities. Yes, yes, and build more role based persona that way. Um departments again, we we already talked about four, you know, teams and departments own their. Catalog items, they own their knowledge. They're the content authors.

SPEAKER_02 22:04

Mm-hmm.

SPEAKER_01 22:04

That clean and tier support paths. So you open a ticket, you need a you know virtual editor flows. It feels like a client-grade interaction, not a help desk form.

SPEAKER_00 22:15

You you've done this a couple of times. I I think you hit upon all the major ones.

unknown 22:20

Yeah.

SPEAKER_01 22:20

So I think those are the four big ones that I would keep in mind when you're trying to bring a firm into a central product.

SPEAKER_00 22:27

Yeah, I don't I don't know if anything to add to that. How do you balance standardization with the autonomy often expected in partnership-driven organizations? I don't know that this applies as much on the portal side, but what does the take?

SPEAKER_01 22:43

Someone without, I mean, when you think of law firms, someone who without without thinking, you would quickly say, well, partners are owners of the firm. They like to run their own show. They don't like to be told what to do, they prefer their own tools, and they have a low tolerance for being told what they have to do. Fine, you're right. But you know what? They're intelligent people, just like any other of your of your stakeholders and users that you're trying to support, they're going to gravitate towards the less friction, easiest way to get the get the work done because this is not what they want to spend their day doing. Attorneys in this in this case, partners want to serve their clients. They don't want to be bundled up in technology or a form or some process within the firm. You know, they they would go to wherever is the quickest way in the most trustworthy way, and probably the most supported way. So that, you know, I think they they go they go to the smart. You've seen this. You know, if you're a tier two or tier three support person and you directly work with an attorney, you know they're going to call you directly from now on and not to help you.

SPEAKER_00 23:43

Yes, that that's that self-learning AI at work.

SPEAKER_01 23:47

Yeah. I mean, that's one. And again, we talked about this before, lead with the friction, you know, not the tools. It's the same thing as we entered before, sort of like uh minimal viable standards.

SPEAKER_00 23:59

And I think to some degree, you can also meet them where they are, providing an omnichannel experience. You have those that prefer to call, those that prefer email, and then you try to integrate that self-service experience on the portal, provide them a form, then they don't have to wait, and then deflection. So I I think giving them options to get what they need, as long as it every option is easy for the way that they're used to working.

SPEAKER_01 24:27

Yeah, I think the last one is CAB, making sure everyone has a seat. It's that give them a seat not a grave, right? So making sure every team has representation in the approval process of changes to their to the environment. So everyone's represented fairly. It's much like our government.

SPEAKER_02 24:44

Well, well, how our government's supposed to work.

SPEAKER_01 24:50

If that happens, they don't resist it as much. They know they're part of team, and they also see their side. They're interacting, right?

SPEAKER_02 24:57

Yeah.

SPEAKER_01 24:57

They understand why things are happening. So I think change a diverse gov diverse cab helps out as well.

SPEAKER_00 25:04

All right. Getting a little bit more technical now.

SPEAKER_01 25:07

All right. I like that. So what reference architecture ensures employees center acts as an experience layer without tightly coupling to underlying domain systems like HRI assembly.

SPEAKER_00 25:21

Yeah. And I think you hit uh upon it really well earlier. I think starting with a taxonomy, starting with where you're going to provide this service and what services you're going to provide on it and thinking holistically about the service is that starting point and where I would start in building out that kind of support architecture. And you're still exposing the different services from this portal. It's just cohesive. If you are a chat person, if you are a search person, multiple ways to interface. If you have chat that is outside, if you use Teams, you can integrate. And I think that would be the starting point.

SPEAKER_01 26:02

Okay, so how do you design that taxonomy so that it's more user intent driven while still mapping to service ownership like SLAs?

SPEAKER_00 26:10

Yeah, about 250 an hour. So this will be different depending on your organization. And assuming you have a product-led model, maybe something a little bit more centralized around decision making, you still have to understand how the business works, who are your stakeholders, who are you building this for? Who is going to use it? And what are the most important services that should be provided there that are one going to help the business do business and just get through the business of being an employee and associate. And then two, drawing a blank one second, more caffeine, more caffeine. And I think part of this starts getting into service modeling where you have tied specific SLAs to specific services or specific requests that you fulfill. So you might say, I'll wait one day for a loner laptop. Maybe it's an emergency. Maybe you can only wait four hours. And so that SLA might be different based off of that need and understanding what the services you're providing, who you're providing it to, where where are they located, and what is that agreement between the customer and IT on how fast that service needs to be provided? Starts to flush out some of that, but you certainly need input from the business, from your stakeholders on what this thing needs to look like. And again, I I thought you covered uh a lot of those things good earlier on. What is your approach to integrating non-ServiceNow systems, SharePoint, Viva, custom apps without degrading the unified experience?

SPEAKER_01 27:52

Well, I think the main thing you gotta keep in mind when you're doing this is don't get don't don't get about the project you're running here because the employee center needs to be the orchestration layer, not the destination for everything. It's not made it shouldn't like some things are just better. Uh workday is better for the HR data in a sense, right? SharePoint is software repositories.

SPEAKER_00 28:15

Might even say this is the front door to the front door.

SPEAKER_01 28:19

It's a skeleton, it's the moving pieces that the transportation, you know, to take you into the experience, right? It's a tour guide, whatever those things are, but it doesn't necessarily have to be the content.

SPEAKER_00 28:32

And I think something I'll add to this is there are very technical solutions on the platform meant to handle these types of situations. So if you do need to bring in data from an external system and it's not native to ServiceNow, first thing you should do is if you have an integration hub, is check your spokes. There probably is already a pre-built integration that you can flip the switch, turn on, depending on which one it is. Not all of them are as easy as flipping the switch, but a good deal are. And that would be the starting point before you go and build custom, before you import spreadsheets, set up schedule jobs. Uh, just use the platform, use spokes and integrate that way.

SPEAKER_01 29:16

Yeah, so I think also you embed it and not have so many hard redirects. You know what I mean? Don't take them somewhere else, you know, make build more embedded experiences that way that seems more seamless, more part of it, that they don't feel like they're being jolted through hyperspace in the sense of some other place and they get car sick or lost. Another thing is probably like Kareem or it's like like what I would do with asset management. So I would use ServiceNow as the orchestration point. So transitions, activity logs, you know, just all the different movements and pieces managing the alert, how they manage the alert, all that content about it. That's what they do well. It's you know, it's incident, problem change, request all those pieces that go around those actions. I would do that in the employee center. So I would have notification patterns, you know, all those pieces, you know, be in the orchestration piece. And I think everything you look at, maybe it comes in one, three buckets, it goes embedded, link to it, or replace it. Because you're that's another probably way of thinking when you're thinking about integrating with third parties, is to fit them in one of three buckets and just see how you how you do it. I mean, you're always gonna get that argument, why can't we just keep the SharePoint site? And maybe they're right. Yeah. I mean, we've got a university site that's a learning LMS to sync it or convert it. They they work well for training. Why can't just tap into it and then have my interface using their content? You know, and so that's anyway. That's to me, I it's a simple approach.

SPEAKER_00 30:54

So passionate about that one, I see. Yeah. How do you prevent portal refragmentation inside employee center through poorly governed topic pages or duplicated content? I I think you you kind of hit on some of that. Um and I I think having that it goes back to governance a little bit when you have changes that are made to the portal. What is the technical governance board reviewing? How is this getting approved? What is the intake mechanism? Who is kind of the demand manager in this case? And when we think about actually, yeah, I so I'd say what is your what is your take on this? I have something I could add to you.

SPEAKER_01 31:35

No, you're right. We cover a lot of it, but I mean it's it's still you have the one front door test for everything. You've got a cab that's represented by all the groups to make to ensuring that it does benefit all the groups and doesn't, you know, slight or deter or become a silo, you know. You just want to fight the, you don't want to win the portal war and then lose the lose the piece. Yeah. Right. So I think those are pieces to that that I think that answers that question.

SPEAKER_00 32:02

All right. AI analysis.

SPEAKER_01 32:04

So where does Gen AI analysis much materially reduce friction versus simply just masking poor information architecture?

SPEAKER_00 32:13

So I think if you there there's this is almost kind of phased. There's parts of AI that you can use without a lot of data. And so it's not really masking it. So I'll give you an example. Incident summarization or a major incident summarization. That is a skill that is using data that's inside the ticket that isn't dependent on your CMDB, your knowledge articles being in a certain format. That is something that you can get value out of right away. That would be a phase one AI use case. And I think that, depending on the use case, the one my favorite is resolution note generation. So you have worked on this ticket, maybe a week, you have added notes to it, you've gotten comments back from the customer, you've added attachments, and you're closing the ticket out. And it is up to you now to summarize everything that you've done to solve that ticket, which will in the future be used to solve it the next time it happens, if it happens the next time. And having those resolution notes generate based off of that data, super valuable, does reduce friction, and that's not masking poor information. And so I'd say anywhere you can use Gen AI as a skill will allow you to achieve that outcome without having too much dependency on the information architecture. When you start getting into re-leveraging data that's using either CMDB data or knowledge article data, and you haven't kind of gone through and given it a second glance or a glow up for AI, that's that would be when you start running into issues with using Gen AI. Skill-based, super valuable.

SPEAKER_01 33:54

So, what governance model do you apply to AI generated responses to ensure of accuracy, compliance, auditability?

SPEAKER_00 34:02

About 250 an hour. No.

SPEAKER_01 34:04

Yeah.

SPEAKER_00 34:07

So yeah, very common. So this is going to vary, and there's no way we're going to answer all of this on this call. I think you start with what you currently have. You start with a technical governance board, you start with your compliance stakeholders, privacy and security to figure out what AI policies do we need to adhere to. And from a privacy perspective, data perspective, what do we need to adhere to? And then from a security perspective, do we have the right RBAC controls? Are these AI agents secured via roles? And do you have to be authenticated to use them? That kind of stuff. Technical governance on the platform, making sure they're doing what they they're supposed to be doing. If you look at different key, I think that's just general. And then when you look at some requirements, this is very common and ServiceNow does this out of the box. Whenever you use an AI capability or anything was generated, it shows you the icon for AI. If AI writes something to the ticket, it says AI wrote this, and then it has a disclaimer for AI. So you're at least aware of what was generated with it. And I think then from a capability standpoint, if you want to formalize or standardize, uh, that's kind of where you start getting into AI control tower and GRC or integrated risk management, where you have those controls and those two kind of work in tandem to help you understand your AI landscape, what models you're using, what prompts, what skills, where they're deployed, uh, and really formalizes and brings structure uh to that. I was ready for that question. You were. So, how do you evaluate whether conversational interfaces will replace or augment structured navigation over the next three to five years?

SPEAKER_01 35:48

So it's augment is my first response, but it will it will shift quickly, and most orgs aren't even ready for the augmentation phase. So I think you got two types of users that you got of. One's a discovery user who's okay with they thrive and conversational, right? And you got the task user, they don't have time for this. They want two clicks, done, get out, you know. So I think you have to structure for both of those types of people. Again, you got to build these things for the people to use them effectively, the roles that they play. And I think that's the only way to be successful. So trust and confidence is another problem, though, you know, to be said, this is what I told me to do. But you know what? I don't think it's right. If I really think of it this way or this way, if the moon was full on a Saturday night in the middle of June, it would work this way.

SPEAKER_00 36:36

You know, sometimes yeah, that that is that kind of gets into evaluations. So regression testing and evaluations on your AI agents against kind of a golden set of data, so that every so often when you're doing your upgrades or the plug-in releases that come out, you're regression testing and making sure that you're still getting those expected responses against data that you know is good and should give you that response.

SPEAKER_01 37:02

And they say also the the interface is like the last thing you should think of. It should be your back end. Otherwise, you might have like a KB database of graveyard of articles and noble entry viewed since the year 2000. Oh, and your conversational AI is want to be real friendly, they give you Darby.

SPEAKER_00 37:23

Microsoft and Carta 2000.

SPEAKER_01 37:26

Yeah, that's what exactly it is. And uh so I think the entry point becomes conversational, but I think you've got to be just protective of both sides of that fence, the task user and the discovery user.

SPEAKER_00 37:37

So to that point right there, uh on the data side, what readiness gaps do you think need to be addressed before AI can deliver meaningful value?

SPEAKER_01 37:47

So what gaps must be addressed? Uh well, knowledge-based quality, that's your sonic pillar.

SPEAKER_00 37:54

Yeah, that's a big one.

SPEAKER_01 37:56

Taxonomy, that's another big one. I don't know, being fresh in life cycle management. I mean, uh I guess adopting, evolving in your life cycles as things change, people change, products change, the way you handle things change.

SPEAKER_00 38:09

I think that's to be there. That goes back to if you have good and a good knowledge base, you can put some of that in there and have AI leverage it and read from it. Hopefully keeping up with process changes as well. I think that uh kind of the the space that doesn't get as much glamour is AI ops or AI operations. So if you have something that breaks, is that service self-healing, or are you calling a server your infrastructure folks who manage the servers at three in the morning because something just needs to be restarted or it's something simple? And I think that coupled with your CMDB data, which has historically always been a gap in some form or fashion, uh, but with AI, with new ways to get data, with all these pre-built integrations, the the data process and the data source issue is starting to go away. And now it it's starting to become a process problem. And I think we can at some point in the future revisit that and then there's gonna be a ton of value there.

SPEAKER_01 39:06

I think feedback is another one. I think it's gotta be more than just ticking a box, thumbs up, thumbs down. You have to point at your phone so you can fix it. And I think you need to keep a good track of the known unknowns issues, you know.

SPEAKER_00 39:21

And track use metrics to track that stuff too. What what skill was used? Who was the user? So you could go solicit some feedback, and what were they trying to accomplish or do? Was it a search and they just left because they couldn't find what they were looking for? Then what AI skills are your most popular? And then where is adoption going down? And what improvement plan can you develop? Yeah. So we're running here, it's the finest turn four, isn't it? Yes, we're we're headed to home base and uh to go from the car and car and garage analogy to baseball. And this is a tune up business case and ROI, the the sweet, sweet numbers at the end that uh hopefully it's not all about. So, how do you quantify the economic impact of portal fragmentation using metrics such as time to resolution, ticket deflection, and employee productivity loss?

SPEAKER_01 40:18

So the first thing I think of it just as one of your friction sale points in the in going with the unified employee center in a sense, is it's not about resolution time only. That's not your cost factor. It's everything before the ticket. When the employee can't find the right portal, they can't they submit to the wrong team, it gets rerouted, they play volleyball with it, or they they give up and they can call the help, you know, they don't call the help desk directly, all those costs you've lost already.

SPEAKER_00 40:47

Yeah, I I love that. That that's was not even thinking about all the things that happened before that ticket submitted. That is a great point. And yeah, I I think it's possible to assign numbers to those activities, track the metric, and then just do the math on that.

SPEAKER_01 41:03

Deflection has some measurements, you know. If you can deflect cleanly, clearly, and correctly, you know Accuracy.

SPEAKER_00 41:11

Accuracy.

SPEAKER_02 41:11

Accuracy, yes. Yes, trying to be waterless. Go ahead.

SPEAKER_00 41:16

I was gonna say, at what scale? So yeah, at what scale does Employee Center Pro become well, I maybe this isn't the the most well-phrased question. I'll ask anyways, and maybe we can adapt. But at what scale does having that front door experience become ROI positive?

SPEAKER_01 41:34

Like how would you think uh it's less about head count, it's more about complexity and discipline in fighting for the the the the the good life for the great the good.

SPEAKER_00 41:44

The good fight, yeah, for attention.

SPEAKER_01 41:47

Yeah.

SPEAKER_00 41:47

I think that's uh well you can probably just figure out deflections, metrics, time savings, try and get some kind of blurry number around that and multiply and from your KPIs. A lot of ways of doing CSAT and surveys and feedback, kind of the the non-quantifiable is also a good thing to try and quantify. Are folks having a better experience getting those testimonials? Uh I think all of that is part of that pie or that picture of how well you those outcomes are being achieved. Yeah. And if you do an app rationalization exercise, add that as well.

SPEAKER_01 42:26

So what fine what what financial model do we use to allocate shared platform costs across departments?

SPEAKER_00 42:33

So this is gonna depend. And I I don't have the answer to this. Um I I have a thoughts on on how we're currently doing it. But you typically have this concept of a peanut butter spread, you have the concept of chargeback. You could do chargeback by sharing costs through a peanut butter spread, you could do chargeback by directly charging back the perceived or actual cost of utilization of that service, or you could just do showback and you don't charge back at all. You just kind of own or eat the cost, but you just show the utilization, you show the business that they're getting value, but it comes out of your budget, which doesn't really feel very good because uh that's like me asking for ice cream and I'm gonna try and think of a different example. But me asking for a nice shiny sports car, I ask you for it and you say, sure, I got nice shiny sports cars. Here, which one do you want? I'll take the Ferrari. Thanks. Oh, and by the way, that'll go on your tab. See ya, yeah, it is a lot what that's like. So to try and make that more fair, if you're gonna provide a service like that out to the greater organization, it should be split. Everybody should have skin in the game, also reduces the costs on whoever would have owned that whole thing, splits it out, and then you also have more resources to deliver more. And I think everybody wants something. So being able to deliver more is not positive.

SPEAKER_01 43:56

And how do you measure the success beyond adoption, especially? Actually, how do you tie that plural usage to business outcomes like retention and efficiencies or even compliance?

SPEAKER_00 44:06

I I think beyond adoption, beyond OCM, so you've got, let's assume you have the utilization. Then it is looking at those KPIs, looking at the OKRs, did you achieve them? Yes or no? And I think that's a way to say that your definition of success has been met. How frequently, okay, so shifting gears one more time, still along the lines of value. So metrics and continuous value. How frequently should executive stakeholders review performance and what metrics actually influence continued investment?

SPEAKER_01 44:42

Well, frequencies, I think like weekly is not your executive, but weekly, I think your platform team looks at it, your service owners look at it, ticket volumes, deflection rates, search abandonments, things like that, especially in the beginning as you're tuning. Um, so that's weekly. Monthly, you want three to five metrics, no narratives, just red, yellow, green. That goes to your executive. And I think quarterly, you want to meet just for some strategic moves. That's when the cab probably moves and meets. Because I think you don't want to meet monthly, you're gonna wear them out. Change fatigue is real. And I think quarterly or biannually is not a bad thing for that cab to make moves, you know, in that in that area.

SPEAKER_00 45:22

Yeah, I'd have nothing to add to that.

SPEAKER_01 45:24

I think what what is it that you're looking at? Costs, displacement, you know, financial, the employee experience, what else?

SPEAKER_00 45:34

Low lowered MTTR, uh deflection being a big one. You probably already mentioned that.

SPEAKER_01 45:40

Yeah, the ROI on your automations for sure.

SPEAKER_00 45:43

Yeah, anywhere where you can show ROI on integrations, interoperability of data, application rationalization, anything that gets either soft dollars back or hard dollars for reallocation is also super beneficial, just above or outside of those typical service desk metrics.

SPEAKER_01 46:03

So, how do you ensure each department sees that measurable value?

SPEAKER_00 46:07

Right at about 250 an hour. What you do, yeah, I'll I'll put it on a brochure. I think one possible way uh or one approach that is helpful is you just have a let me think about this. I think you can provide a success dashboard, essentially, which shows stats, statistics, have it selectable by departments or organization. Maybe you don't have departments and you have a BU uh or BUs, show it at that level. I think there has to be OCM in some champ champion enablements in order for this to be successful. If you just put a dashboard up, no one's gonna look at it. Everybody's busy. But if you are able to get a message in front of somebody, point them at a dashboard or just a screenshot from that dashboard, link to it, send that out in an email, have a share it out on a company call, at least three different ways to show that value. Because if you only choose one way to do it, you're gonna miss two-thirds of folks.

SPEAKER_01 47:04

So yeah, I think if you had three to five, seven, maybe at the most, thing that you said kills our organizational effectiveness with these several silo portals in day one, and you have these, and this is our metrics today.

SPEAKER_02 47:18

And you show the metrics as you progress and how that those numbers are decreasing or are eliminated. Yeah. Yeah.

SPEAKER_00 47:30

And hopefully you've got some CSAT and the testimonials to go with that.

SPEAKER_02 47:35

Yeah.

SPEAKER_00 47:36

But yeah, that's kind of showing that execution against those outcomes. And even better if your sponsor, either your executive sponsor, can speak to it as well.

SPEAKER_01 47:46

You know, when we talk about this topic, I thought it'd be about a 30, 40 minute podcast. We're about 15 minutes.

SPEAKER_00 47:51

This was this was stump the chump, uh, and these were Yeah. These are very difficult questions.

SPEAKER_01 47:58

Yeah. All right, so we're pulling it back in the garage and shutting it down for the day. So here's where I want everyone to walk away with. A single door isn't just a service now project. It's not technical, it's not technology. Um, it's a it's a commitment, though. It's a commitment to your employees and that they that when they need something, they don't have to go hunting. The technology is ready, now assist is ready, the platform can carry the weight. The question is whether your organization is ready and ready to have that hard conversation about ownership, governance, and who controls what and why. So if if you're in a in a business in a regulated industry or any environment where trust and confidentiality aren't optional, that single door has to be built right, not just fast. So, Steele, as always, great conversation today. Thank you for supporting in this and bringing us together. If this episode hits home for you, do us a favor, please. Share with somebody who needs to hear it, and leave us a review, subscribe, subscribe, subscribe. We need subscribers.

SPEAKER_02 48:57

That's for the algorithm.

SPEAKER_01 48:58

Yeah, come find us on YouTube, help us out, please. We're dropping content every week for the practitioners, the product owners, and the leaders who are out there in the trenches doing the real work. I'm Pops, and I'll keep Frenching.

SPEAKER_00 49:10

Cheers.